The OpenStreetMap (OSM) community is constantly expanding. Over 1.5million individuals have contributed data to it as per the stats page on the OSM Wiki. This network expands every day to include people with a wide range of interests and professions. There are numerous tools, technologies, and approaches being created to facilitate the contribution and use of OSM data. Even though there are forums, local social media networks, and wiki pages that can help users learn more about OSM, someone who is unfamiliar with the community and ecosystem may not be able to navigate the resources effectively. Trying to find relevant answers, switching pages, and waiting hours for forum responses can be a daunting task.
To address these issues, we present an OSM chatbot which comes to the rescue. Chatbots are easily one of the most well-known examples of artificial intelligence (AI). But, if you are new to the concept of a chatbot, it is an AI-based computer program that simulates human conversations. Chatbots are also known as digital assistants that understand human capabilities. Bots interpret the user intent, process their requests, and give prompt relevant answers. They can communicate through voice as well as text and can be deployed across websites, applications, and messaging channels such as Facebook Messenger, Twitter, or WhatsApp. Even in Nepal, we can see chatbots being used as virtual assistants in banking software and other apps.
Kathmandu Living Labs, as a pioneer of OpenStreetMap in Nepal, is constantly working on innovative ideas and tools to benefit the OSM community around the world. As a result, we developed this chatbot with two major aims:
- Assist new users with basic OSM queries and introduce them to the community.
- Bringing all of the OSM tools together in one place. For example, displaying user contributions, locating tags, displaying events related to OSM, and forwarding to certain wiki pages and forums.
- Provides smart & instant response to OSM-related queries.
- Gives the most relevant tag information about a specific feature.
- Displays OSM contribution stats made by users.
- Provides ease of search by displaying the relevant recommendations.
- Offers user suggestion mechanism to improve the answer.
- Obviously has 24X7 online support.
- Integrated with messaging channels. ( Currently integrated on Facebook, but can be easily integrated across other channels like telegram, slack, etc. )
Fig: A working example of the OSM chatbot
We gathered data on queries and responses from a variety of sources, including the OSM wiki, forums, and pilot testing performed within the Kathmandu Living Labs team. Currently, the chatbot is in the early stages of development and can only respond to a few OSM-related questions. We hope to improve the chatbot’s performance in the coming days. We believe that the more users interact with the chatbot, the better the results will be. So, we expect the community to engage with our project by interacting with the chatbot, providing comments, or assisting us with technical issues.
There are a number of future improvements we hope to add to this project- in the short and medium-term:
- Support for multiple languages.
- Increase the training data set to support a wide range of user queries.
- Scale the computational resources.
These selected tasks, however, are preliminary. We hope to solicit feedback and incorporate them in refining the chatbot as more OSM community members use the tool. We believe that as we learn and expand as a community, this chatbot will become a global platform for individuals interested in learning about OSM, as well as a starter kit for growing OSM-based tools and companies in the chatbot world.
You can ask queries related to OpenStreetMap at facebook.com/osmchatbot.